Self-regulated learning (SRL) is a well-known concept within the field of online learning; however, tracking, interpreting, and scaffolding effective SRL strategies can be challenging. SRL encompasses complex and dynamic processes involved in monitoring and controlling the cognitive, motivational, and behavioral components of learning. Such processes include task interpretation, application of knowledge and strategies to solving that task, and monitoring the relative success of implementing a particular strategy. Metacognitive monitoring is a critical component of effective implementation of SRL. Without accurate monitoring, a learner doesn't have adequate information to know whether and how to adjust their learning choices. This chapter discusses two studies that explore how feedback valence influences learners' metacognitive perceptions of their performance and how those (mis)perceptions impact both their expectations for future success as well as SRL choices for a subsequent task.
TopIntroduction
Self-regulated learning (SRL) is considered key to student success across a wide range of student learning contexts (e.g., Dent & Koenka, 2016; Donker, De Boer, Kostons, Van Ewijk, & van der Werf, 2014; Pintrich, 2000). SRL is a broad conceptual term used to encompass the multivariate cognitive, affective, and motivational processes that underlie students’ dynamic learning behaviors. Key models of SRL (see e.g., Panadero, 2017 for an overview) emphasize the importance of distinct SRL cognitive and affective mechanisms and behaviors such as metacognitive processes, goal setting, and emotion and motivation regulation. In general, however, there is agreement across frameworks that the learner is an active participant in the strategic control of their own knowledge and skills and that SRL has an important impact on student achievement.
Much of the educational research on SRL has examined the relationship between students’ SRL capacities and strategies with their achievement outcomes (e.g., Dent & Koenka, 2016), and on the efficacy of a variety of SRL training interventions (e.g., Jansen, Van Leeuwen, Janssen, Jak & Kester, 2019) with metacognitive processes often a key construct of interest. Like SRL, metacognition can be broadly construed—encompassing numerous regulatory behaviors that rely on monitoring and control of current and prior knowledge, adaptive deployment of learning and planning strategies, and goal setting and revision. The deployment of the constellation of metacognitive processes that are involved in SRL show individual differences that relate to academic performance (e.g., Winne, 1996).
From a cognitive perspective, metacognitive evaluations of what one knows, of recent performance outcomes, or of the current rate of learning are directly linked to metacognitive control processes (e.g., Nelson & Narens, 1990). Because metacognitive monitoring directly influences regulatory behaviors (e.g., Metcalfe & Finn, 2008), the accuracy of those evaluations is critical. The accuracy of metacognitive monitoring relates to the kinds of cues that people use to evaluate their knowing (e.g., Koriat, 1997). Monitoring can be subject to systematic biases when learners rely on cues that are not diagnostic of actual learning (e.g., Finn & Tauber, 2015).
The learning context has an important influence on the cues that are available for learners to evaluate their ongoing cognition. For example, when information is presented in such a way that it feels more fluently processed (e.g., large versus small font, louder versus quieter) people evaluate that their learning as better than when it feels less fluently processed even though perceptual fluency doesn’t impact actual learning. (e.g., Murphy, Huckins, Rhodes & Castel, 2022). We present two studies that manipulated features of the design of the learning task to investigate how they would influence students’ metacognition about their performance and as a result their regulatory choices about the difficulty of their next task.
We chose to investigate feedback valence and autonomy as our design features of interest because as learning has increasingly become more digitally mediated in the classroom and in online and blended learning environments, educators and assessment developers have been utilizing games as a medium to support learning and assessment. Feedback valence and autonomy are two commonly used game features. The gamification of learning or assessment involves designing or adding game elements (such as points, feedback, interactive mechanics, and graphics) into a digital assessment or a learning task. Gamification is used in educational contexts because it is thought to increase student motivation and engagement (e.g., Gee, 2003; Prensky, 2001; Shute & Ke, 2012), which may positively benefit learning and assessment performance. Though some research has explored the impact of gamification in learning and instruction (e.g., Kapp, 2012) there is considerably less research on how gamification influences assessment behaviors and performance. In the current study we explored how two different features—feedback valence and autonomy-- influenced students’ performance, metacognitive evaluations of their own performance, and choices about task difficulty on a subsequent task.